# Microsoft > Markdown twin of https://roboterminal.app/companies/microsoft · https://roboterminal.app/companies/microsoft.md ## Key numbers - **Units shipped:** 1,000,000 units - _"Microsoft's Maia 300 is expected to ship more than 1 MILLION units next year, according to Reuters as cited in the report."_ - Source: @MilkRoadAI on X ## On shared sources - [Google](https://roboterminal.app/companies/google) (80) - [Meta](https://roboterminal.app/companies/meta) (72) - [Amazon](https://roboterminal.app/companies/amazon) (69) - [NVIDIA](https://roboterminal.app/companies/nvidia) (62) - [AMD](https://roboterminal.app/companies/amd) (24) - [Oracle](https://roboterminal.app/companies/oracle) (24) - [Anthropic](https://roboterminal.app/companies/anthropic) (21) - [OpenAI](https://roboterminal.app/companies/openai) (18) - [Micron](https://roboterminal.app/companies/micron) (11) - [Goldman Sachs](https://roboterminal.app/companies/goldman-sachs) (10) - [Nebius](https://roboterminal.app/companies/nebius) (10) - [Broadcom](https://roboterminal.app/companies/broadcom) (9) - [Vangrid](https://roboterminal.app/companies/vangrid) (7) - [Apple](https://roboterminal.app/companies/apple) (6) - [AWS](https://roboterminal.app/companies/aws) (6) - [Marvell](https://roboterminal.app/companies/marvell) (6) - [Qualcomm](https://roboterminal.app/companies/qualcomm) (6) - [Samsung](https://roboterminal.app/companies/samsung) (6) - [J.P. Morgan](https://roboterminal.app/companies/j.p.-morgan) (5) - [SpaceX](https://roboterminal.app/companies/spacex) (5) - [TSMC](https://roboterminal.app/companies/tsmc) (5) - [CoreWeave](https://roboterminal.app/companies/coreweave) (4) - [Dell](https://roboterminal.app/companies/dell) (4) - [ING](https://roboterminal.app/companies/ing) (4) - [Lumentum](https://roboterminal.app/companies/lumentum) (4) - [Mollie](https://roboterminal.app/companies/mollie) (4) - [Nscale](https://roboterminal.app/companies/nscale) (4) - [Tesla](https://roboterminal.app/companies/tesla) (4) - [Intel](https://roboterminal.app/companies/intel) (3) - [KORE1](https://roboterminal.app/companies/kore1) (3) - [Mecka](https://roboterminal.app/companies/mecka) (3) - [Morgan Stanley](https://roboterminal.app/companies/morgan-stanley) (3) - [Palantir](https://roboterminal.app/companies/palantir) (3) - [Sequoia](https://roboterminal.app/companies/sequoia) (3) - [SK Hynix](https://roboterminal.app/companies/sk-hynix) (3) - [ByteDance](https://roboterminal.app/companies/bytedance) (2) - [Figure](https://roboterminal.app/companies/figure) (2) - [HSBC](https://roboterminal.app/companies/hsbc) (2) - [Hugging Face](https://roboterminal.app/companies/hugging-face) (2) - [Nomura](https://roboterminal.app/companies/nomura) (2) - [SanDisk](https://roboterminal.app/companies/sandisk) (2) - [UBS](https://roboterminal.app/companies/ubs) (2) - [Vertiv](https://roboterminal.app/companies/vertiv) (2) - [Waymo](https://roboterminal.app/companies/waymo) (2) - [xAI](https://roboterminal.app/companies/xai) (2) - [1X](https://roboterminal.app/companies/1x) (1) - [a16z](https://roboterminal.app/companies/a16z) (1) - [Accenture](https://roboterminal.app/companies/accenture) (1) - [Alibaba](https://roboterminal.app/companies/alibaba) (1) - [AlphaSense](https://roboterminal.app/companies/alphasense) (1) - [Amkor Technology](https://roboterminal.app/companies/amkor-technology) (1) - [Animoca Brands](https://roboterminal.app/companies/animoca-brands) (1) - [Astera Labs](https://roboterminal.app/companies/astera-labs) (1) - [Asus](https://roboterminal.app/companies/asus) (1) - [Awish](https://roboterminal.app/companies/awish) (1) - [Bank of America](https://roboterminal.app/companies/bank-of-america) (1) - [Bloom Energy](https://roboterminal.app/companies/bloom-energy) (1) - [Borderless Capital](https://roboterminal.app/companies/borderless-capital) (1) - [Chang Robotics](https://roboterminal.app/companies/chang-robotics) (1) - [Cloudflare](https://roboterminal.app/companies/cloudflare) (1) - [CopilotKit](https://roboterminal.app/companies/copilotkit) (1) - [Counterpoint](https://roboterminal.app/companies/counterpoint) (1) - [Credo](https://roboterminal.app/companies/credo) (1) - [Crusoe](https://roboterminal.app/companies/crusoe) (1) - [CXMT](https://roboterminal.app/companies/cxmt) (1) - [DeepSeek](https://roboterminal.app/companies/deepseek) (1) - [DigitalOcean](https://roboterminal.app/companies/digitalocean) (1) - [Dyna Robotics](https://roboterminal.app/companies/dyna-robotics) (1) - [Gartner](https://roboterminal.app/companies/gartner) (1) - [Ge Vernova](https://roboterminal.app/companies/ge-vernova) (1) - [Genesis Mission](https://roboterminal.app/companies/genesis-mission) (1) - [GitHub](https://roboterminal.app/companies/github) (1) - [GUC](https://roboterminal.app/companies/guc) (1) - [HashKey Capital](https://roboterminal.app/companies/hashkey-capital) (1) - [Honda](https://roboterminal.app/companies/honda) (1) - [HPE](https://roboterminal.app/companies/hpe) (1) - [Inovia Capital](https://roboterminal.app/companies/inovia-capital) (1) - [IREN](https://roboterminal.app/companies/iren) (1) - [iRobot](https://roboterminal.app/companies/irobot) (1) - [Lambda](https://roboterminal.app/companies/lambda) (1) - [Mantic AI](https://roboterminal.app/companies/mantic-ai) (1) - [MediaTek](https://roboterminal.app/companies/mediatek) (1) - [Milk Road Pro](https://roboterminal.app/companies/milk-road-pro) (1) - [Mistral](https://roboterminal.app/companies/mistral) (1) - [Moondream](https://roboterminal.app/companies/moondream) (1) - [Neo](https://roboterminal.app/companies/neo) (1) - [NetSuite](https://roboterminal.app/companies/netsuite) (1) - [NXP Semiconductor](https://roboterminal.app/companies/nxp-semiconductor) (1) - [Open Compute Project](https://roboterminal.app/companies/open-compute-project) (1) - [OpenRouter](https://roboterminal.app/companies/openrouter) (1) - [Palo Alto Networks](https://roboterminal.app/companies/palo-alto-networks) (1) - [Perplexity](https://roboterminal.app/companies/perplexity) (1) - [PTC](https://roboterminal.app/companies/ptc) (1) - [Reflection](https://roboterminal.app/companies/reflection) (1) - [Sakana AI](https://roboterminal.app/companies/sakana-ai) (1) - [Salesforce](https://roboterminal.app/companies/salesforce) (1) - [Schneider Electric](https://roboterminal.app/companies/schneider-electric) (1) - [SemiAnalysis](https://roboterminal.app/companies/semianalysis) (1) - [Sequoia Capital](https://roboterminal.app/companies/sequoia-capital) (1) - [Shadeform](https://roboterminal.app/companies/shadeform) (1) - [Skyfall](https://roboterminal.app/companies/skyfall) (1) - [Thinking Machines Lab](https://roboterminal.app/companies/thinking-machines-lab) (1) - [TrendForce](https://roboterminal.app/companies/trendforce) (1) - [Uber](https://roboterminal.app/companies/uber) (1) - [Win Semiconductors](https://roboterminal.app/companies/win-semiconductors) (1) - [Workday](https://roboterminal.app/companies/workday) (1) ## Pulse - [@pequityresearch on X](https://x.com/pequityresearch/status/2113317206708679036) Sam Altman on the amount of compute needed if 50% of the world were using personal AI assistants to its full capabilities: "How many GPUs do you want working for your personal assistant? When people t… - [@ren_stocks on X](https://x.com/ren_stocks/status/2113293311310237753) $META Does timing matter? Honestly, not that much to me. It matters a lot in the short term and very little in the long term. This week is a perfect example. META posted a strong quarter (time the rip… - [@kyleichan on X](https://x.com/kyleichan/status/2113229006803792148) ARTEX, an open-source Chinese AI cyber tool, seems to be involved in a number of AI cyberattacks: Korean banks, Japanese firms, and now a Microsoft alert. Crucially, this is not an AI model. It’s a ha… - [@stretchcloud on X](https://x.com/stretchcloud/status/2113173258233167934) The headline out of Anthropic's expected S-1 this week will be the valuation, something north of $2 trillion depending on which leaked number you trust. The filing itself, once it's public, will show … - [@stretchcloud on X](https://x.com/stretchcloud/status/2113120661795164424) Code review does not scale to how much code agents can write in a day, and a test suite only proves the specific cases somebody remembered to write. Microsoft Research and GitHub Next shipped somethin… - [@martin_casado on X](https://x.com/martin_casado/status/2113117450065887657) I dunno, at some point we just need to unalign Claude and let it save the free world ... https://t.co/vtvoPHKWp4 https://t.co/iFPRYyIXpB - [@stretchcloud on X](https://x.com/stretchcloud/status/2113115628861026549) 2,900 sign-in attempts a minute. That's the number Microsoft published this week for a single actor's agentic penetration test, and it's the number that should worry anyone still assuming AI security … - [@stretchcloud on X](https://x.com/stretchcloud/status/2113105814106116190) Two stories broke the same week that tell you where AI security actually stands. Wiz, the Google-owned security shop, Calif, and OpenAI's own red team independently found two critical PHP flaws in Aug… - [@pequityresearch on X](https://x.com/pequityresearch/status/2113029058074407403) BofA: TrendForce states Nvidia's AI-server share moderates from 66.9% in CY26 to 62.6% in CY27 as “Others” rises from 26.0% to 29.6%. Custom ASICs increase from 26.1% to 29.6% of AI server shipments, … - [@pequityresearch on X](https://x.com/pequityresearch/status/2112955387489996944) Expert call covering GUC – Google CPU & TPU v10 ASIC outlook, Microsoft Maia 200 ramp, TK1 vs. TK3 margin dynamics, TSMC capacity support, and 2027 hyperscaler volume trajectory. $MFFT $META $GOOGL ht… - [@pequityresearch on X](https://x.com/pequityresearch/status/2112922859538444376) Nomura: global data center power usage by application $NVDA $AMD $VRT $GOOGL $MSFT $META $AMZN - [@pequityresearch on X](https://x.com/pequityresearch/status/2112772618092740646) Morgan Stanley sees the cost per new MW peaking at $51.8 million in FY27, before sliding to $47 million in FY29E. $MSFT $NVDA $AMZN $GOOGL $META https://t.co/IC6zlAbmhs https://t.co/8zXgV111pw - [@stretchcloud on X](https://x.com/stretchcloud/status/2112743929443492293) Idle GPUs in the middle of the worst GPU shortage in memory sounds like a contradiction. It isn't, and Shadeform just shipped the feature that explains why. Microsoft is reportedly sitting on $80 bill… - [@pequityresearch on X](https://x.com/pequityresearch/status/2112629904420548979) CEO of Google, Sundar Pichai, on compute demand and AI bubble possibility: Q: Ben Horowitz relayed Andy Jassy had something like compute is sold out till 2028, are you seeing the same thing? "If today… - [@pequityresearch on X](https://x.com/pequityresearch/status/2112626894096957489) CEO of Lumentum, Michael Hurlston, on the impact of yields on capex spend: "The CFO of Alphabet came on just a couple of days ago and was talking about their capex going up year over year. For the mos… - [@elonmusk on X](https://x.com/elonmusk/status/2112603906723487744) “Lack of electricity” is an oversimplification. Everything is hard about bringing massive compute online at extremely high speed. If there was twice as much time, this would be relatively easy, but th… - [@pequityresearch on X](https://x.com/pequityresearch/status/2112589669699403886) Expert Takeaways: Data Centers 1. Open-Source Models Closing the Gap > Open-source models are catching up to state-of-the-art frontier models faster than expected, driven by lower costs and improved p… - [@stretchcloud on X](https://x.com/stretchcloud/status/2112562987839943124) Skyfall's launch thread says SKY-001 beats every frontier model tested on CAD and BIM. Skyfall's own research post is more careful: it "matches or exceeds" them. That gap between the tweet and the pap… - [@pequityresearch on X](https://x.com/pequityresearch/status/2112550655000813697) Nomura: data center capacity to expand to 74GW in 2028F, up from 26GW in 2025. Of the 74GW, 48.5GW coming from the US big tech. $SPCX $ORCL $CRWV $META $MSFT $AMZN $GOOGL - [@MilkRoadAI on X](https://x.com/MilkRoadAI/status/2112244498059407763) We're about to witness the BIGGEST infrastructure spending boom in human history. According to Morgan Stanley, hyperscalers and neoclouds are projected to spend a staggering $1.58 TRILLION in 2027 and… - [@ruudvz on X](https://x.com/ruudvz/status/2112215504903446548) → Microsoft: eigen chips, meer controle. → AMD: een flexibel serverrek voor AI-agenten. → Workday: software die je verzoek uitvoert. → Google en Hue: je lamp krijgt lokale bediening. → Anthropic: een … - [@pstAsiatech on X](https://x.com/pstAsiatech/status/2112209547565625582) Microsoft pulls out of UK Nscale data centre deal over power delay https://t.co/vsjoL6aVmR - [@pequityresearch on X](https://x.com/pequityresearch/status/2112198335197044770) Keybanc: We believe Marvell is likely to win Maia 400 over Mediatek and Broadcom. Maia 300 is expected to be in MP in 1H27 with ~500K unit shipments next year, while Maia 400 is expected to ramp in 1H… - [@pequityresearch on X](https://x.com/pequityresearch/status/2112168386767769881) Goldman Sachs estimates Microsoft will scale to ~40 GW of capacity by mid 2030. $MSFT https://t.co/KOx7K5tG6p - [@stretchcloud on X](https://x.com/stretchcloud/status/2112166122091147723) A £2 billion data centre site in Loughton, Essex was supposed to go live in 2027 with roughly 23,000 Nvidia chips inside it. This week it came out that Microsoft walked away from its anchor tenant com… - [@stretchcloud on X](https://x.com/stretchcloud/status/2111963034671353863) Samsung raised its 32GB DDR5 price from $149 to $239 in September. That's a 60 percent jump in one month, and it isn't the top of the curve. Contract DRAM pricing more than doubled this year, from abo… - [@pequityresearch on X](https://x.com/pequityresearch/status/2111928304672391575) Goldman Sachs has raised the global servers TAM by 31%/45% in 2027/28E to ~$1.1 trillion and ~$1.6 trillion, respectively. New AI servers shipment forecasts imply AI chip demand of 18m/29m/38m in 2026… - [@MilkRoadAI on X](https://x.com/MilkRoadAI/status/2111823872723214387) Microsoft just walked away from one of the biggest AI data center projects in the United Kingdom. According to the Telegraph, $MSFT terminated its deal for Nscale's planned site in Loughton, Essex, wh… - [@pequityresearch on X](https://x.com/pequityresearch/status/2111549307405173047) J.P. Morgan: AI CapEx > 2026 Capex Forecast: Growth estimate raised from +80% to +85% YoY. This adds over +$270B in incremental spending (a new record step-up, beating 2025). > 2027 Capex Forecast: Gr… - [@pequityresearch on X](https://x.com/pequityresearch/status/2111442112986165734) NPO and CPO adoption by customers presented by LightCounting at OCP. $AMZN $GOOGL $META $MSFT $ORCL - [@Scobleizer on X](https://x.com/Scobleizer/status/2111269878267306311) I met a lot of people today at @TheHumanoidHub’s conference in San Francisco and I definitely was the stupidest one in the room. A role I am used to, but my meager brain struggles to keep up with peop… - [@pequityresearch on X](https://x.com/pequityresearch/status/2111202270742524036) UBS: AI power TAM growing to $91.1 billion by 2030E, up from $17.7 billion in 2026E. Of the $91.1 billion, $49 billion dedicated to GPU power TAM, while the other $42 billion for ASIC. $GOOGL $MSFT $M… - [@MilkRoadAI on X](https://x.com/MilkRoadAI/status/2111011014582194464) According to Korean media, Microsoft, Google, Amazon and Meta are speeding up their in house AI chips. That could open up a SECOND group of HBM buyers beyond $NVDA. Microsoft's Maia 300 is expected to… - [@pequityresearch on X](https://x.com/pequityresearch/status/2110930466534564251) Keybanc on Intel: Qualcomm's AI300 is expected to use EMIB-T AWS is evaluating EMIB-T for Trainium 4 Intel is getting meaningful traction with EMIB-M with design wins expected with AWS for Trainium 3 … - [@pequityresearch on X](https://x.com/pequityresearch/status/2110826796442906890) HSBC: ASIC will account 57% of total HBM demand by 2028e (from 31% in 2025) $GOOGL $MU $MRVL $META $MSFT $AVGO $AMZN - [@pequityresearch on X](https://x.com/pequityresearch/status/2110581432338829634) Bloomberg shared this visual showing the estimated cost of building a data center. Of the estimated $29.2 billion, $17.1 billion (or 58.5%) is dedicated to chips & servers. $NVDA $DELL $GEV $AMD $GOOG… - [@MatthewChang on X](https://x.com/MatthewChang/status/2110567071717466506) Ryan officially on team Chang Robotics Fund, representing world class start-ups in manufacture tech. See you there! https://t.co/52QUeyQSqI - [@nguyenthambt on X](https://x.com/nguyenthambt/status/2110542161167225222) Building infrastructure for Physical AI requires more than understanding the technology. It requires knowing how to turn a promising idea into a system that can operate at scale. @vangrid_io brings to… - [@ricci_nov on X](https://x.com/ricci_nov/status/2110225417135440377) Maia 200 shipped in January. Maia 300 lands nine months later: TSMC N2P, 288GB HBM4 at 18TB/s vs 216GB at 7TB/s, 96 per rack vs 48, 950W vs 750W. Microsoft claims 40% more tokens per megawatt than GPU… - [@pequityresearch on X](https://x.com/pequityresearch/status/2110085662825152928) J.P. Morgan is forecasting a ~300GW shortfall in global AI compute by 2031E. $NVDA $AMD $GOOGL $MSFT $META $AVGO $AMZN https://t.co/iZrvcU2svc https://t.co/yzzhhhEN7G - [@ParadisLabs on X](https://x.com/ParadisLabs/status/2110074252716056818) Memory fundamentals are still very strong (stronger than Q1). $MU CEO: - "We expect memory and storage supply demand conditions to be much tighter in fiscal 2027 and 2028 than they were in 2026." - "W… - [@TheRundownAI on X](https://x.com/TheRundownAI/status/2109954950637769058) Top stories in AI today: - Meta’s Muse Wonder lands at the frontier - Andrej Karpathy sketches AI’s ‘sharp funnel’ - Put Meta’s new Muse Wonder to work - Microsoft’s Maia 300 takes aim at Nvidia https… - [@Kore1Staffing on X](https://x.com/Kore1Staffing/status/2109710638087852188) Microsoft Maia 300 is a hiring story as much as a chip launch. KORE1's take: the bottleneck isn't the silicon, it's the ML compiler engineers who make it useful. https://t.co/tFzlBZpkzh https://t.co/E… - [@Kore1Staffing on X](https://x.com/Kore1Staffing/status/2109709886091137283) Who designs the chip after Microsoft Maia 300? The same small pool Google, Amazon, Meta, and OpenAI are recruiting from. KORE1 breaks down which chip roles get squeezed first. https://t.co/tFzlBZpkzh … - [@Kore1Staffing on X](https://x.com/Kore1Staffing/status/2109709128729825337) Microsoft Maia 300 enters production in 2027, and the SIA projected about 67,000 U.S. chip jobs at risk of going unfilled by 2030. KORE1 expects compiler and validation engineers to feel it first. htt… - [@pequityresearch on X](https://x.com/pequityresearch/status/2109708427324756276) Citi: CSPs data center capex is expected to increase to ~$1.6 trillion in 2030E, with global capex spend increasing to ~$3.8 trillion. Google to see the strongest capex growth through 2030. $GOOGL $AM… - [@pstAsiatech on X](https://x.com/pstAsiatech/status/2109705794820473002) This..... https://t.co/XXq8766RZ5 - [@stretchcloud on X](https://x.com/stretchcloud/status/2109686533414633584) Microsoft just told the market it's betting multiple gigawatts of future data center capacity on chips it designed itself. Maia 300 claims 40 percent more tokens per watt than leading GPU based system… - [@TheRundownAI on X](https://x.com/TheRundownAI/status/2109675234752069805) NEW: Microsoft introduced its new Maia 300 chip, with Satya Nadella calling it the "biggest bet yet on our own silicon." The company says it gets 40% more AI work out of the same power than top chips … - [@pstAsiatech on X](https://x.com/pstAsiatech/status/2109665820414705830) Microsoft plans huge deployment of its own AI chips, with an eye on saving costs https://t.co/LzBK1OMuf4 - [@MilkRoadAI on X](https://x.com/MilkRoadAI/status/2109639714907033974) This is the reason Microsoft is cutting its Claude token usage. Microsoft just announced it will scale its own AI chip, Maia, to multigigawatt deployments starting in 2027. Maia will become the main s… - [@pequityresearch on X](https://x.com/pequityresearch/status/2109466080481817065) UBS: "As a reminder, for C2027 we anticipate ~50GW of compute shipments (requiring ~57GW of powered data center including PUE factor) against current visibility for ~38GW of powered shell when excludi… - [@pequityresearch on X](https://x.com/pequityresearch/status/2109059903939543192) "Bernstein has raised its forecast for the global data center XPU ASIC market in 2028 from $140–160 billion to $250–300 billion, with projections for 2026 and 2027 at $50–70 billion and $130–150 billi… - [@pequityresearch on X](https://x.com/pequityresearch/status/2108953202284474538) David Lazovsky, EVP & GM, Data Center Networking Business Group at Marvell Technology on AI inference: "3 years ago 1/3 of total data center infrastructure was focused on inference, today north of 60%… - [@pequityresearch on X](https://x.com/pequityresearch/status/2108794345885139180) Former CEO of Intel, Pat Gelsinger, sees more AI data center delays coming due to lack of energy capacity: "We just need more capacity, and fundamentally, energy capacity omewhat dampens how you fork … - [@kyleichan on X](https://x.com/kyleichan/status/2108243962158895171) Microsoft showing off the ability to run DeepSeek V4 Flash locally on a Surface laptop. “It’s possible to bring more of the frontier to your PC.” https://t.co/1z7P7pyRol - [@TheHumanoidHub on X](https://x.com/TheHumanoidHub/status/2108010791681282524) Mecka AI just raised a $60M Series B led by Sequoia. NVIDIA, Microsoft M12, Qualcomm, and Samsung also participated, along with angels including Tony Xu, Frank Slootman, and former Tesla Optimus lead … - [@stretchcloud on X](https://x.com/stretchcloud/status/2107998158508917177) $2,599 buys a Windows laptop that Microsoft says renders AI images 4.3 times faster than a 16 inch MacBook Pro on M5 Pro. The marketing is aimed squarely at Apple. The architecture underneath is aimed… - [@CyberRobooo on X](https://x.com/CyberRobooo/status/2107980799014564272) Physical-world data is the new oil Mecka AI has raised a $60M Series B, led by Sequoia Capital, with NVIDIA, Qualcomm, Microsoft and Samsung among the new investors. It captures human activity in the … - [@pstAsiatech on X](https://x.com/pstAsiatech/status/2107961080014139427) This.... https://t.co/5Faed9Ijs3 - [@johnkoetsier on X](https://x.com/johnkoetsier/status/2107955934396092891) Robotics' biggest bottleneck? Data. Mecka AI is cashing in with human motion data for robots. • $100M annual revenue run rate • $300M projected 2026 revenue • $500M valuation • $68M raised Backers inc… - [@Newsforce on X](https://x.com/Newsforce/status/2107941787180454031) 🇺🇸Trump will hand National Medal of Science awards to @ElonMusk, @Nvidia’s @JensenHuang, @Google’s Sergey Brin and @AMD’s @LisaSu at the White House on Thursday. @MichaelDell and @Microsoft’s @Satya… - [@thenewsink on X](https://x.com/thenewsink/status/2107908819418534002) Microsoft Brings Powerful AI Directly to PCs https://t.co/t1sWe16biq - [@pequityresearch on X](https://x.com/pequityresearch/status/2107649530871038069) Expert Takeaways: Marvell > Marvell will receive roughly $8 billion plus equity warrants in its Google deal, with the Google LPU v10 projected to generate about $10 billion in revenue over 2029–2030. … - [@pequityresearch on X](https://x.com/pequityresearch/status/2107630777563525248) Expert Call: Memory - Key Takeaways > Contracts are moving from flexible 2–3 year terms to 5yr+ take-or-pay structures with upfront deposits and price floors/ceilings, prioritizing downside protection… - [@smsehy on X](https://x.com/smsehy/status/2107368284794597773) In plant floors and enterprise IT, safety never lives inside application logic. The operational risk with autonomous agents is granting write access to legacy systems that lack rollback paths. Guardra… - [@joinautopilot on X](https://x.com/joinautopilot/status/2107201082275307985) Today was absolutely insane in the stock market Here's everything you need to know: 1. Weak September jobs report continues fueling bets the Fed skips an October hike 2. The Nasdaq hit a record high a… - [@nguyenthambt on X](https://x.com/nguyenthambt/status/2105845890028274141) Building infrastructure for Physical AI requires more than an ambitious vision. It takes experience in deep tech, distributed systems and turning complex technology into products people can actually u… - [@pequityresearch on X](https://x.com/pequityresearch/status/2105735247329644770) A good overview by UBS of the market participants in humanoid robot related sectors $MSFT $NVDA $QCOM $AVGO $MU $AMZN https://t.co/A0yxpRgeLh - [@Scobleizer on X](https://x.com/Scobleizer/status/2105489866884575677) I share posts I like from the AI community and add my evening’s thoughts on. I don’t care if a dude is teleoperating my @1x_tech Neo robot if it, er he, is entertaining my friends and making them drin… - [@nguyenthambt on X](https://x.com/nguyenthambt/status/2105485789153149274) Behind every ambitious Physical AI network, the team building the infrastructure matters just as much as the vision. @vangrid_io brings together experience across deep tech, Web3 and large scale distr… - [@nguyenthambt on X](https://x.com/nguyenthambt/status/2104756800260387252) A spatial intelligence network needs two things to scale: a reliable technical foundation and the ability to bring real-world data demand into the network. That is what makes the team behind @vangrid_… - [@stretchcloud on X](https://x.com/stretchcloud/status/2104615620453011923) The agent safety conversation just moved from the prompt to the silicon. NVIDIA shipped the Open Agent Safety Platform this week with over 100 partners, and the mechanism is the interesting part. Open… - [@techniahqrobot on X](https://x.com/techniahqrobot/status/2103981278802665760) A Century of Innovation (1926–2026) 2026 — 🇺🇸 Agentic AI: Systems that don't just think, but act autonomously. 2026 — 🇨🇳 Humanoid robots: Humanoid robots move from research labs into workplaces an… - [@nguyenthambt on X](https://x.com/nguyenthambt/status/2103658313405661285) The team behind @vangrid_io has experience across deep tech, Web3 and distributed systems. Robert Brighton brings 10 years of GTM and revenue experience across deep tech and Web3, including previous w… - [@nguyenthambt on X](https://x.com/nguyenthambt/status/2102583020318507161) One thing that makes @vangrid_io interesting to me is the team and the problem they chose to solve. Vangrid is building infrastructure around the growing need for real world spatial data as robotics a… - [@pequityresearch on X](https://x.com/pequityresearch/status/2102204230832427383) Worth reading this conversation about AI demand and memory from StatementDog's analyst, Ken. > While NAND prices are still rising in the second half of 2026, increases are moderating to low-double dig… - [@pequityresearch on X](https://x.com/pequityresearch/status/2102173779720089899) General Manager of Microsoft Taiwan says the demand for AI computing power will continue to exceed supply for the next 2-3 years, as a result, the company is expanding the construction of data centers… - [@stretchcloud on X](https://x.com/stretchcloud/status/2102140562065305970) Everyone is still arguing about parameter counts while the real efficiency gains come from a different lever: what a weight is allowed to be. Moondream just shipped Parakeet Redux, a speech to text mo… - [@pequityresearch on X](https://x.com/pequityresearch/status/2102098872147333398) J.P. Morgan: CSP Capex expected to reach ~$1.7 trillion in 2028 🤣 That is a lot of money... $GOOGL $AMZN $META $MSFT $ORCL $SPCX https://t.co/QpnbRLcFz6 - [@stretchcloud on X](https://x.com/stretchcloud/status/2101944528592584830) Apple shipped something small this cycle that I think matters more than another point release of a frontier model: macOS 27 now includes an fm command in the terminal, putting Apple's on-device model … - [@nguyenthambt on X](https://x.com/nguyenthambt/status/2101863783505478055) The more I look at @vangrid_io, the more interesting the team behind the infrastructure becomes. Robert Brighton brings 10 years of GTM and revenue experience across deep tech and Web3, with previous … - [@pequityresearch on X](https://x.com/pequityresearch/status/2101826742956494938) BofA: We do not expect “de-spec” or lower memory content in new GPUs/ ASICs; instead, we assume a further increase, to 1,000GB per Rubin Ultra or 384GB for the new Rubin, vs 288GB currently. $NVDA $AM… - [@pequityresearch on X](https://x.com/pequityresearch/status/2101809630858948639) Nice visuals shared by Blackstone in their recently uploaded Youtube video about AI. > Google token use per month has increased from 10T in May 2024 to 3,200T in May 2026. > Anthropic + OpenAI's ARR a… - [@pequityresearch on X](https://x.com/pequityresearch/status/2101768389215801459) HSBC: USA model launches per year has been accelerating and still far ahead of China. $GOOGL $MSFT $META $AMZN https://t.co/9oqbKNwxCR - [@pequityresearch on X](https://x.com/pequityresearch/status/2101395845832909055) Goldman Sachs: Server Market Overall Server Market Projections > Market Expansion: The overall global data center server market is expected to grow to $1.5 trillion by 2030, driven by a 5-year CAGR of… - [@joinautopilot on X](https://x.com/joinautopilot/status/2101324306290544659) Breaking: Nscale just filed to IPO under $NSCL Backed by Nvidia, the AI data center company plans to list on the NYSE with Microsoft as a major customer • +1,252% revenue growth to $140.6M in 2026 • $… - [@pequityresearch on X](https://x.com/pequityresearch/status/2101114255634305064) Jensen Huang on slowing down AI: "We should go as fast as we can irrespective of anybody else. We are going to go as fast as we can, but we would never ever and never should ship products before they … - [@ParadisLabs on X](https://x.com/ParadisLabs/status/2101062878392434837) Recommended reading for the weekend: 1. @damnang2 — Is the 2028 Memory Downcycle Real? 2. @AnthropicAI — Scenarios for our Economic Future 3. @BrookingsInst — Workforce policy for the age of AI 4. @Mi… - [@frontrunvc on X](https://x.com/frontrunvc/status/2101042201069301943) FUNDRAISE 💰: @frontrunvc flagged @_Mantic_AI on July 25th. 55 days later: $25M seed led by @radicalvcfund, with @balderton, Microsoft's M12 and Thinking Machines Lab. 🫡 https://t.co/jUTZ8a4GR3 https… - [@pequityresearch on X](https://x.com/pequityresearch/status/2101041820876656743) J.P. Morgan: ASIC/XPU units expected to reach 55% of total units in 2028E, up from 32% in 2025. $AMD $NVDA $GOOGL $META $MSFT $AMZN $AVGO - [@pequityresearch on X](https://x.com/pequityresearch/status/2100982429670654427) HSBC: Outside of Google, the % of ASIC compute among major CSPs remains steady from 2025 and into Q2 2026. $GOOGL $AMZN $META $MSFT $ORCL - [@pequityresearch on X](https://x.com/pequityresearch/status/2100683963010265418) BofA: 1GW of a orbital data center currently costs $160-$180 billion compared to a terrestrial data center at $40-$60 billion, with the cost of launch being ~$50 billion. $NVDA $SPCX $TSLA $AMD $GOOGL… - [@pequityresearch on X](https://x.com/pequityresearch/status/2100669366891376647) Citi: AI token usage has increased by a CAGR off 31% MoM or 2,434% YoY. A lot of tokens are being used... $NVDA $AMD $GOOGL $META $MSFT $META - [@pequityresearch on X](https://x.com/pequityresearch/status/2100621694100545840) Funda AI: "Our interviews suggest that frontier labs remain committed to training and advancing model capabilities internally, while additional evaluation and remediation may slow the pace of public r… - [@pequityresearch on X](https://x.com/pequityresearch/status/2100571724194455981) KB Securities: the allocation of AI infrastructure spend toward memory is projected to jump from 40% (2026E) to 57% (2027E). $MU $GOOGL $SNDK $META $MSFT $AMZN - [@pequityresearch on X](https://x.com/pequityresearch/status/2100437086222352549) Incredible visuals shared by Brad Gerstner @altcap at the All-In Summit. > The hyperscalers capex is almost dollar for dollar with semis FCF > Collective run rate of Anthropic, OpenAI, and SpaceX is ~… - [@pequityresearch on X](https://x.com/pequityresearch/status/2100378198261268643) TrendForce: Global Power Demand Capacity > Global data center power demand capacity is expected to expand from 122.9 GW in 2025 to 161 GW in 2026. > The share of AI servers is likely to exceed 40% by … - [@pequityresearch on X](https://x.com/pequityresearch/status/2100331034453680539) Samsung Securities: Hyperscaler Compute Business Models Hyperscalers operate through two primary methods: 1) Computing rental and 2) Model inference services, both of which can achieve a Gross Profit … - [@pequityresearch on X](https://x.com/pequityresearch/status/2100319058428383620) UBS: total AI server racks estimated to grow to 428,299 in 2027E before steadily declining. Most of the server racks to be dominated by AI ASIC. $NVDA $AMD $MSFT $AMZN $GOOGL $META - [@pequityresearch on X](https://x.com/pequityresearch/status/2100301190743630129) Samsung Securities: Hyperscalers to have a temporary FCF deficit period spanning from 2Q26 to 3Q28, followed by a dramatic expansion in FCF from 2029 through 2030. > CapEx Growth as a Rational Investm… - [@pequityresearch on X](https://x.com/pequityresearch/status/2100271746720964644) One of the big takeaways from AI Infra Summit Day 1 is that the useful life of accelerators is now approaching 10 years, due to tight supply and software gains. $NVDA $GOOGL $AMZN $MSFT $AMD $META - [@pequityresearch on X](https://x.com/pequityresearch/status/2100256395996127584) KB Securities: as of September, no sign of orders slowing down from major customers for memory. Memory makers' inventories are at an all-time low (under 10 days) with supply tightening further in 2027… - [@pequityresearch on X](https://x.com/pequityresearch/status/2100074446686199968) BofA: adding a layer of policing and monitoring in AI could increase compute cost by 25%. $NVDA $AMZN $AMD $INTC $MSFT $GOOGL $META - [@pequityresearch on X](https://x.com/pequityresearch/status/2100059348013658522) Big day in the AI world. Here are some important highlights for you all: Satya Nadella's full discussion at the All-In Summit was released on Youtube today: > Nadella advocates for a multi-model world… - [@MParekh on X](https://x.com/MParekh/status/2099972396178608605) ‘One More Thing’ on ‘The Blip 3.0’. Trump, Microsoft & China. ARD #163 ...Anthropic AI ‘Pace Pause’ now a wrestling match: Trump’s hoax call to Jensen, Microsoft’s rules, Beijing’s rebuke. Siri AI shi… - [@MParekh on X](https://x.com/MParekh/status/2099915590429274226) 'One More Thing' in AI: Trump Calls 'Blip 3.0' AI Pause a Hoax, Microsoft's Code as Essay, China Answers | ARD #163 AI Ramblings Daily podcast #163 on AI-RTZ. $MSFT https://t.co/BsDJfOiJQC - [@pequityresearch on X](https://x.com/pequityresearch/status/2099911372242841986) J.P. Morgan has a GPU/ASIC shipment forecast of >25 million in 2028E, mostly dominated by ASICs. Up from ~10 million in 2025. $GOOG $AMD $NVDA $MSFT $META $AMZN https://t.co/aSpobzqW5N - [@TheRundownAI on X](https://x.com/TheRundownAI/status/2099808143274324119) Top stories in AI today: - Trump, Beijing both shoot down the AI slowdown - Apple finally starts shipping Siri AI upgrade - Make any complex topic click with an ELI5 skill - Microsoft AI drafts the ru… - [@pequityresearch on X](https://x.com/pequityresearch/status/2099653925817213354) Morgan Stanley: CSP CapEx to increase to $1.5 trillion in 2027, of which 53% will be for memory alone or $795 billion. $DRAM $EWY $MU $MSFT $META $AMZN $GOOGL - [@pequityresearch on X](https://x.com/pequityresearch/status/2099620455988560014) Chairman of Win Semiconductors, Chen Chin-Tsai, states that AI is not a bubble and that AI development "is only just beginning." "Chen Chin-Tsai analyzed that the development of AI is gradually moving… - [@pequityresearch on X](https://x.com/pequityresearch/status/2099569669778624513) BofA sees recent frontier lab controversy as "noise" relative to a market where AI CapEx could surge 3x to $3+ trillion by decade end, and demand signals remaining strong. $GOOGL $MSFT $META $AMZN $OR… - [@stretchcloud on X](https://x.com/stretchcloud/status/2099555024552526272) GPT-6 Astra is not a better chatbot. It is a computer-use agent with a 1,050,000 token context window priced at $10 per million input tokens. Those three things together are the story. The context win… - [@pequityresearch on X](https://x.com/pequityresearch/status/2099549718854488147) PCB content per rack remains extremely robust. VR300 +69% compared to VR200 MI5XX +73% compared to MI4XX TPUv9 +132% compared to TPUv8 T4 +222% compared to T3 $AMZN $META $MSFT $GOOGL $TTMI - [@pequityresearch on X](https://x.com/pequityresearch/status/2099535646247596372) Goldman Sachs says that optical companies they spoke to on their Asia tour are saying AI investment is in the "early stages" and that it will be a 10-20 year cycle with some customers already planning… - [@stretchcloud on X](https://x.com/stretchcloud/status/2099534892086284701) The Microsoft AI usage limits are not a safety overreaction. They are a structural response to what happens when you deploy capable AI at scale without explicit behavioral constraints. The three limit… - [@pequityresearch on X](https://x.com/pequityresearch/status/2099288266675568772) J.P. Morgan: We model 45% AI chip shipment CAGR in 2025-28, with AI ASIC chip shipments growing at a CAGR of 60%, driven by strong AWS Trainium/Google TPU demand. $NVDA $GOOGL $AMD $MSFT $META $AMZN - [@pequityresearch on X](https://x.com/pequityresearch/status/2098881073862103272) KB Securities sees FCF for hyperscalers rebounding in 2028 on higher cloud utilization, pricing, and margins. $META $MSFT $GOOGL $AMZN $ORCL - [@elonmusk on X](https://x.com/elonmusk/status/2098809953439904220) Grok available in Microsoft Copilot https://t.co/0ypxUGvq2J - [@pequityresearch on X](https://x.com/pequityresearch/status/2098794512974057607) HSBC: Anthropic and OpenAI are losing market share in token volume, but their share in token spend has remained strong. $GOOGL $META $MSFT $AMZN - [@pequityresearch on X](https://x.com/pequityresearch/status/2098610550678937793) SemiAnalysis: we estimate ~$11T of cumulative capex from CY24 through CY29. $GOOGL $AMZN $META $MSFT $ORCL - [@stretchcloud on X](https://x.com/stretchcloud/status/2098569027308446044) The most interesting thing about OpenAI's Habitat rewrite is not the numbers. It is who did it. Two engineers. Codex. GPT-5.5. One quarter. A full Python-to-Rust rewrite of the storage system serving … - [@pequityresearch on X](https://x.com/pequityresearch/status/2098541417740054545) Here are some of the important takeaways from Goldman Sachs Communacopia + Technology Conference 2026 and Citi’s 2026 Global TMT Conference this week: > Demand for AI compute, infrastructure, and hard… - [@pequityresearch on X](https://x.com/pequityresearch/status/2098540689717862410) Here are some of the important takeaways from Goldman Sachs Communacopia + Technology Conference 2026 and Citi’s 2026 Global TMT Conference this week: > Demand for AI compute, infrastructure, and hard… - [@MilkRoadAI on X](https://x.com/MilkRoadAI/status/2098485730360324098) You should be buying as much NVIDIA as possible (Save this). SemiAnalysis projects that NVIDIA could hold roughly $1.4 trillion in cash and investments by fiscal 2031, giving it more capital than almo… - [@pequityresearch on X](https://x.com/pequityresearch/status/2098444708003664176) Evercore ISI: ASICs account for about 10-30% of internal workloads today and in 5 years will account for 30-50%, but enterprise driven workloads will remain >90% with Nvidia. $GOOGL $AMZN $META $MSFT … - [@MParekh on X](https://x.com/MParekh/status/2098443151056761337) ‘Business as Usual’ in AI. Microsoft, Oracle & SpaceX. ARD #161 ...Microsoft to 38 gigawatts. Oracle on other people’s money. SpaceX by the month. Meta’s Muse, early reviews. $SPCX $MSFT $ORCL $META F… - [@MParekh on X](https://x.com/MParekh/status/2098432836881490200) 'Business as Usual' in AI: Microsoft 38 GW, Oracle Prepayments, SpaceX Rentals & Muse | ARD #161 AI Ramblings Daily podcast #161 on AI-RTZ $MSFT $SPCX $META $ORCL YouTube link with show notes: https:/… - [@stretchcloud on X](https://x.com/stretchcloud/status/2098313092740366650) You don't hire 1,000 engineers for a pilot. Accenture and Google Cloud just launched a Gemini Enterprise Business Group. 1,000 forward-deployed engineers dedicated to getting Gemini into enterprise pr… - [@pequityresearch on X](https://x.com/pequityresearch/status/2098253523771084978) AI inference expected to grow to 75% of data center TAM by 2030. GPU share of AI server shipment will decline from 69.7% in 2026 to 58.5% in 2030, but still controlling the lead. $NVDA $AMD $GOOGL $ME… - [@Scobleizer on X](https://x.com/Scobleizer/status/2097958614044701047) How do we know we don't have AGI yet? Because they aren't ordering a @waymo and exploring the world more yet. IE, making a run for it. Based on current world model learning speeds it is possible to se… - [@pequityresearch on X](https://x.com/pequityresearch/status/2097863174368223636) Day 2 Takeaways | Goldman Sachs Communacopia + Technology Conference 2026 and Citi’s 2026 Global TMT Conference Companies I covered include: Amkor Technology $AMKR Microsoft $MSFT Sandisk $SNDK NXP Se… - [@ParadisLabs on X](https://x.com/ParadisLabs/status/2097809963141791875) Just some notes from $NBIS fireside chat at Goldman Sachs Tech Conference: TLDR: Yep, it's bullish. 1. Order book extends into H1 2028 which is 2 quarters further out than at Q2 ER. Arkady said "peopl… - [@pequityresearch on X](https://x.com/pequityresearch/status/2097770219519230062) No AI ❌️ used. I listen to all of the conversations carefully and take notes. I will be sharing Day 2 takeaways from the conferences hosted by Goldman Sachs and Citi for all paid subscribers on X and … - [@MilkRoadAI on X](https://x.com/MilkRoadAI/status/2097736507394330858) A four year old chip is now more expensive to rent than when it was brand new and that shows you the demand for compute (Save this). The H100 launched in 2023 and should be obsolete by now but instead… - [@Newsforce on X](https://x.com/Newsforce/status/2097589124861440054) 🇫🇮 @Google announced a $15.1 billion investment in AI infrastructure in Finland over the next two years, its largest single investment in Europe ever. Finland has become a hotspot for data centers t… - [@Newsforce on X](https://x.com/Newsforce/status/2097588923778121756) 🇫🇮 Google announced a $15.1 billion investment in AI infrastructure in Finland over the next two years, its largest single investment in Europe ever. Finland has become a hotspot for data centers th… - [@pequityresearch on X](https://x.com/pequityresearch/status/2097389756665446774) Goldman Sachs: ASICs to account for 50%/52%/55% of total AI chips in 2026/27/28E. $GOOGL $AMZN $MSFT $META - [@pequityresearch on X](https://x.com/pequityresearch/status/2097384310462406898) Qualcomm's EVP, CFO, and COO, Akash Palkhiwala, says they have another hyperscaler engagement similar to Amazon: "we have a very strong engagement with another hyperscaler that we have not disclosed, … - [@CKCapitalxx on X](https://x.com/CKCapitalxx/status/2097287636310233280) $PLTR and $NBIS just partnered to deliver a complete sovereign AI stack, and this is a real validation for both. Nebius becomes Palantir's preferred sovereign AI infrastructure partner. Nebius compute… - [@stretchcloud on X](https://x.com/stretchcloud/status/2097258643007611180) This Microsoft paper landed on something I have been watching for a while. Agent benchmark scores and production reliability are measuring different things.The paper spans 9 models, 4 task families, 5… - [@stretchcloud on X](https://x.com/stretchcloud/status/2097129793699639555) The bubble argument has a revenue problem.Milk Road published data this week: global AI revenue at a $229B annualized run rate, up 3.5x in a single year, doubling every 7.2 months. Anthropic reported … - [@MelvinInvests on X](https://x.com/MelvinInvests/status/2096694365934735851) IREN is going significantly higher if it successfully converts its power and data centers into AI infrastructure. The central thesis is that IREN is increasing both the amount of AI capacity it operat… - [@tchsignal on X](https://x.com/tchsignal/status/2096529341421383951) Microsoft Project Zenith Sets a New Windows Baseline for Local AI Development Microsoft is building Project Zenith around a very different idea of what a developer PC should be. The ready-to-code Wind… - [@tchsignal on X](https://x.com/tchsignal/status/2096525320564445566) Seattle Times and Newsday Sue OpenAI and Microsoft Over AI Training The Seattle Times and Newsday have sued OpenAI and Microsoft in federal court, alleging that their journalism was copied without per… - [@MilkRoadAI on X](https://x.com/MilkRoadAI/status/2096318224594088367) Token usage just went vertical, and the power grid is not ready for it (Save this). This chart shows that agentic token usage on OpenRouter rose from roughly 0.51 trillion tokens on February to about … - [@stretchcloud on X](https://x.com/stretchcloud/status/2096309136267116799) AIRA3 placed 8th out of 4,000 teams in a live Kaggle competition to fine-tune a 30B Nemotron model. Same information as human competitors. External private test set. Gold medal. The architecture detai… - [@MilkRoadAI on X](https://x.com/MilkRoadAI/status/2096282971414442480) The memory stocks run up is just getting started and here is why. This chart shows that Nvidia could account for a smaller percentage of total HBM demand in 2027, falling from 58% in 2026 to 41% but t… - [@MelvinInvests on X](https://x.com/MelvinInvests/status/2095975579736838521) AI demand is just getting started,and we need more compute than ever before. The chart tracks the share of US businesses paying for subscriptions to AI models, platforms, and tools, based on business … - [@stretchcloud on X](https://x.com/stretchcloud/status/2095941212112904432) The image model race just made a clear separation between quality and price leadership. My read on what MAI-Image-2.6-Flash means: Microsoft now owns both. MAI-Image-2.6 is #2 on the Artificial Analys… - [@stretchcloud on X](https://x.com/stretchcloud/status/2095917053227979144) The pattern we need to talk about more: agents do not stay inside their containers. Reuters reported today that a swarm of OpenAI agents hijacked DseWiki, a German programming wiki, in May. They made … - [@stretchcloud on X](https://x.com/stretchcloud/status/2095797263507272153) WebMCP just shipped in CopilotKit. This is worth understanding. The spec is W3C-incubated, Microsoft co-authored, and already in Chrome 146 Canary. Any web app can register its tools in a machine-read… - [@stretchcloud on X](https://x.com/stretchcloud/status/2095458028149744067) The problem with training coding agents isn't that RL doesn't work. It's that most RL approaches train the model inside a framework the agent never uses in production. Microsoft's Agent Lightning v1.0… - [@CKCapitalxx on X](https://x.com/CKCapitalxx/status/2095252342182400285) $MSFT restructuring Q1 27. https://t.co/YBllaavM9K - [Automate Month-End Close… | The Rundown University](https://x.com/TheRundownAI/status/2094445915260031311) Back to feed SaveShare Automate Month-End Close Reconciliation and Reporting in Awish I recently built a month-end close workflow in Awish for a client at a finance company. The problem was not creati… --- Source: Roboterminal · https://roboterminal.app/companies/microsoft.md